Cool Joule GTi
Notes: same Run battery, different Charge battery was meant at 01:40
Figures don’t tally, as had written while doing the run, corrected below.
100uF cap across the power inputs reduces current by approx 45uA.
Without it, the light is slightly brighter and output potential rises to around 1.35V, but charging is seemingly no quicker. The output is higher than the Charge battery voltage anyway with the capacitor in place.
Checking the figures of the first run again – i’d written it on the paper during the run.
14mV were used, not 9mV.
Charge battery went up by 360mV not 349mV.
The Run battery recovered by 5mV for the second run.
After overnight resting, the Charge battery of the first run settled to 1.214V.
The best way of running is to remove the top diode in the circuit diagram.
RimstarOrg pointed out the natural easier path and after much more testing, it is indeed best to leave out that diode. The problem has been that many transistors will quit when that diode is removed. That diode also allows low uA LED running on one battery (the one on the left of the circuit, with the other omitted).
The transistor which changed the circuit again for performance has been the C2236.
It uses 1.9mA, but the circuit runs in a GTi Turbo mode !
Figures below are for the beginning of tests yesterday, with very well rested batteries.
Run 1.306V Charge 0.954V
Late last night, I tried the C2236 and here are the start and end voltages by 10.30am
Run 1.294V Charge 1.063V
Run 1.235V Charge 1.346V
LED output comparatively very bright, half full brightness.
After a swap over run, the batts will be well rested for final voltages.
